One of the great things about living in a railroad town is the stories one gets to hear from the locals. One of our customers at the cafe was a manic artist-type, who at one point had purchased a private railroad car made by the Harriman company that had derailed some time back and had sat on the bank of the Upper Sacramento River for a number of years. Harriman cars are distinctive for their concrete chassis' which give them a smooth ride and greater resistance to the usual side to side rocking of lighter private cars like Pullmans.
I once asked one of my other customers, a railroad employee, how the car came to end up down the hillside.
Now, it should be mentioned here that the Dunsmuir yard is not flat exactly, but on a slight grade. Railroads generally construct their lines with very gradual climbs and descents, so the motive power can be easily regulated. On account of Dunsmuir's grade, locomotive engines are left running so the airbrakes can remain pressurized. Cars on the other hand, when disconnected from the engine and thus the air pressure, have to be "tied down" by hand. Handbrakes are operated by either a pump handle connected to a chain via a ratchet, or a steering wheel looking thing that operates the clamp on the brake. The Dunsmuir grade begins its ascent below Lake Shasta, over fifty miles away.
The Harriman car was sitting near the roundhouse when the brakeman noticed that it was moving slightly. He jumped up on the end of the car to tighten the handbrake and noticed that the brake on the Harriman car was on the inside of the car. This was usually not a problem, but this time the door happened to be locked. I suspect that because the car was elegantly furnished and privately owned, the brakeman decided it would be better to stop the car from the outside rather than smash open a window to get at the brake. Perhaps he would try to stop the car and find a key.
In any case, the brakeman jumped off the slowly rolling car and decided to chuck the wheels of the car. This trick worked with the usual boxcars in the yard. However, the concrete base base of the Harriman car gave it more weight than a boxcar, and thus more force. A bunch of maintenance of way guys looked up from their lunches to see the poor brakeman tossing wood and wheel chucks under the car, which either crushed or pushed aside the objects he threw in front of it.
Once the car moved through the first switch, the men left their lunches and quickly drove to the end of the yard, about a mile downstream. By the time the men got there, the car was already approaching a slow and steady rate of speed. "Rate of acceleration" is probably more accurate, since that was exactly what was happening. As the Harriman car approached them, the men quickly piled in the truck and raced to the next road crossing, at Soda Creek, five miles downstream.
The men jammed in all sorts scrap metal into the tracks to derail the car at Soda Creek. They pounded in tie plates and stacked railroad ties; they constructed a small but sturdy berm with sledgehammers and shovels. moving as much as they could before the car arrived. When they heard the telltale squeal of the trucks moving on the curve, they rain behind the truck and waited.
The Harriman car hit the pile of debris at approximately fifty miles per hour, about at least twice as fast as the men had ever seen anything move along that stretch of track. The Harriman's concrete base gave it a low center of gravity, which now kept it glued to the tracks at speeds that would have derailed all but the lowest profile rolling stock. It did not appear to be slowed in the slightest by the pile which it distributed the instant of the collision.
A second crew, which had been shadowing the runaway, sped past Soda Creek and decided to get off at the Gibson siding, another ten miles downstream to throw the switch to the sidiing. They quickly jumped out of the truck and sent the fastest runner to throw the switch, to get the car off the mainline.
The car hit the switch and the siding at approximately seventy miles per hour, and for a moment, it looked as though it would go through the exit switch and continue on its way to Redding. Instead, when it hit the second switch it crossed over the mainline, and flew the air. It plowed down cottonwoods, oaks, and cedars as it landed upright, continuing in the same general direction it had been heading for the past 15 miles. The car came to a rest with a terrible crash some 200 feet downstream from the switch.
The concrete base had done all the work of clearing its path, and so the car came to a stop pretty much undamaged. Moving the car presented an interesting challenge. When boxcars derail, as they often do in that particular canyon, they are unloaded (and if possible, salvaged) before the car is brought back up. The Harriman car carried its weight in its construction, and could not be thus lightened. Eventually, the railroad disconnected the trucks, which were undamaged and could be reused. Then, much later, the car itself was lifted out of the canyon and onto a flatbed car.